OBJECTIVE: We investigated illness beliefs of recently hospitalized patients with coronary artery disease (CAD) and the prospective association between these beliefs and adherence to secondary prevention behaviors. Causal attributions of CAD and their concordance with actual patient risk profiles were also examined. METHOD: A prospective study of 193 patients was conducted. Data were collected by self-report and from medical records at 3, 6, and 9 months after discharge. Baseline depression was assessed by structured clinical interview. The association between illness beliefs and adherence was tested with hierarchical linear regression controlling for clinical and demographic confounders. RESULTS: Most participants perceived high personal and treatment control and believed CAD to be chronic in duration with severe consequences. A relatively low number of symptoms were endorsed as being part of CAD. Heredity was considered the single most important and most commonly perceived cause of CAD. Smoking, alcohol, emotional state, and heredity were significantly more likely to be endorsed as causal factors by respondents with these risk profiles. In multivariate analysis, illness beliefs contributed an additional 6% of the total variance explained by the model (p = .02). Perceptions of more serious consequences predicted better adherence (p = .03). Social desirability was the best single predictor of adherence. CONCLUSION: Patient perceptions of risk factors were largely consistent with actual risk factors. Despite modest effect sizes, illness beliefs do contribute to our understanding of adherence to secondary prevention behavior. Interventions aimed at modifying these beliefs, particularly those related to the consequences of CAD, may improve patient outcomes.